From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx4-phx2.redhat.com (mx4-phx2.redhat.com [209.132.183.25]) by dpdk.org (Postfix) with ESMTP id 69A57374F for ; Wed, 21 Sep 2016 07:11:29 +0200 (CEST) Received: from zmail09.collab.prod.int.phx2.redhat.com (zmail09.collab.prod.int.phx2.redhat.com [10.5.83.11]) by mx4-phx2.redhat.com (8.13.8/8.13.8) with ESMTP id u8L5BRLw026593; Wed, 21 Sep 2016 01:11:27 -0400 Date: Wed, 21 Sep 2016 01:11:27 -0400 (EDT) From: Panu Matilainen To: Bruce Richardson Cc: Adrien Mazarguil , dev@dpdk.org Message-ID: <1547360604.1032267.1474434687150.JavaMail.zimbra@redhat.com> In-Reply-To: <20160920155848.GA74952@bricha3-MOBL3> References: <1474295814-4879-1-git-send-email-bruce.richardson@intel.com> <20160919145959.GB17252@6wind.com> <20160919152605.GG45548@bricha3-MOBL3> <20160920125126.GG17252@6wind.com> <20160920155848.GA74952@bricha3-MOBL3> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [86.60.134.172, 10.5.101.181] X-Mailer: Zimbra 8.0.6_GA_5922 (ZimbraWebClient - FF48 (Linux)/8.0.6_GA_5922) Thread-Topic: net/mlx: fix compile errors with ignore pedantic pragma Thread-Index: KxVbkfkNG6ZmsnvImlkN3Ce+qdm+JQ== Subject: Re: [dpdk-dev] [PATCH] net/mlx: fix compile errors with ignore pedantic pragma X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 21 Sep 2016 05:11:29 -0000 ----- Original Message ----- > On Tue, Sep 20, 2016 at 02:51:27PM +0200, Adrien Mazarguil wrote: > > On Mon, Sep 19, 2016 at 04:26:05PM +0100, Bruce Richardson wrote: > > > On Mon, Sep 19, 2016 at 04:59:59PM +0200, Adrien Mazarguil wrote: > > > > Hi Bruce, > > > >=20 > > > > On Mon, Sep 19, 2016 at 03:36:54PM +0100, Bruce Richardson wrote: > > > > > With recent gcc versions, e.g. gcc 6.1, compilation of mlx driver= s > > > > > with > > > > > debug enabled produces lots of errors complaining that "pedantic"= is > > > > > not a warning level that can be ignored. > > > > >=20 > > > > > error: =E2=80=98-pedantic=E2=80=99 is not an option that controls= warnings > > > > > [-Werror=3Dpragmas] > > > > > #pragma GCC diagnostic ignored "-pedantic" > > > > > ^~~~~~~~~~~ > > > > >=20 > > > > > These errors can be removed by changing the "-pedantic" to > > > > > "-Wpedantic". > > > >=20 > > > > Nice to have a workaround, I thought they did not keep the option a= t > > > > all. > > > > However after testing: > > > >=20 > > > > - It does not seem to work with GCC 4.6 and older, they prefer > > > > -pedantic: > > > > "warning: unknown option after `#pragma GCC diagnostic' kind". > > > >=20 > > > > - GCC 4.9 (possibly 5.x as well) does not care, can use either > > > > -pedantic or > > > > -Wpedantic. > > > >=20 > > > > - GCC 6 can only supports -Wpedantic. > > > >=20 > > > > Note we're working toward removing the need for these #pragma in th= e > > > > first > > > > place as soon as possible, however in the meantime I fear that chec= king > > > > the > > > > GCC version is necessary. > > > >=20 > > > Depends on how old of GCC version we need to support. From the releas= e > > > notes > > > it appears that -Wpedantic was introduced in GCC 4.8 (3 1/2 years ago= ). > > >=20 > > > https://gcc.gnu.org/gcc-4.8/changes.html > > >=20 > > > Do we need to support compilation on gcc versions older than this? > >=20 > > I'm all for upgrading so I do not really mind if we stop caring about o= lder > > GCC versions (especially considering this problem only occurs in debugg= ing > > mode which is seldom used by non-developers). The version check is > > necessary > > if we want to keep full compatibility with at least: > >=20 > > - RHEL <=3D 6.x > > - Debian <=3D 7.x > > - Ubuntu <=3D 13.04 > >=20 > > Works for me either way, thus: > >=20 > > Acked-by: Adrien Mazarguil > >=20 > Any objections to dropping of support for debug settings for these OS's? No objections on dropping RHEL <=3D 6 support, we never did DPDK on those o= ld versions anyway. As for the others, I've no particular opinion but certainly no objections either. - Panu - >=20 > /Bruce >=20